DOES THE DIURNAL ACTIVITY PATTERN OF CARABID BEETLES DEPEND ON SEASON , GROUND TEMPERATURE AND HABITAT ?

The influence of season, ground temperature and habitat on diurnal epigeic activity of ground beetles (Coleoptera: Carabidae) in floodplain forest and neighboring clear-cut area was studied in late spring and early autumn by pitfall trapping. Among the material collected were 35 species of ground beetles. We recorded a significant influence of daytime factors on epigeic activity in 16 species. The epigeic activity of 13 species was co-affected by the temperature of the soil surface. The activity of some species differed, depending on season and locality.


INTRODUCTION
Based on their diurnal activity, carabid beetles are usually divided into three categories -strictly diurnal species, species with nocturnal activity, flexible species with activity modified by microclimatic and geographic conditions (Thiele, 1977, Novák, 1977).However, we can find differences in activity among not only species and populations of the same species, but even within a single population (Thiele and Weber, 1968).Species can also show various preferences to light intensity in different phases of ontogenetic development (differing periods of development of larva and imago -Carabus problematicus Herbst) or changes in periodicity of reproduction (Carabus cancellatus Illiger).
Annual rhythms are also closely connected to the circadian rhythm -90% of species reproducing in autumn are active at night, while among species reproducing in spring the proportion of nocturnal activity is substantially lower, about 33% (Thiele and Weber, 1968).
Diurnal activity is primarily regulated by endogenous physiological factors, so-called timers, but the length of its duration could be modified by environmental conditions.Light is known to be the most important factor influencing carabid beetle activity (Thiele, 1977), nevertheless temperature and moisture levels can also play an important role in relation to the type of habitat (Thiele, 1977, Novák, 1980, Kegel, 1990).
Studies dealing with diurnal activity can be conducted under natural or laboratory conditions.In laboratory studies, direct observations or instruments recording activity on the weight basis used to be employed (Park, 1935).Actographs are the instruments used to record activity on the basis of electric impulses, infra-red light, or sound (Park, 1935;Backlund and Ekeroot, 1950).Novák (1978) designed a special trap with automatic sampling for laboratory research.In natural conditions, two basic tools are commonly used: 1) simple pitfall traps in various modifications (Williams, 1958, Loreau, 1986), or 2) automatic traps that separate catch by individual periods (Novák, 1979(Novák, , 1980)).

MATERIALS AND METHODS
The study area was located in Litovelské Pomoraví Protected Landscape Area (Central Moravia, Czech Republic) with floodplain forest and meadows along the Morava River.We compared two localities -old floodplain forest (Querco-Ulmetum) and neighboring clear-cut (49°65´N, 17°20´E, altitude 210 m a.s.l.).
The herbal layer of the floodplain forest consists of Anemone nemorosa, Polygonatum spp., Lathyrus vernus and Maianthenum bifolium; the dominant moss was Eurhynchium hians.In November 1998, litter biomass (dry weight) was 622 g/m 2 .The alluvial soil is loamy-sandy to loamy at the locality, with pH 4.8-5.Annual precipitation was about 520 mm; the mean annual temperature was 9.1°C.Part of the forest was cut in November 2002 and replanted in March 2003 by oak, elm and lime tree (ratio 8:1:1) using heavy forestry machines.Before replanting, the remaining wood was chipped and scattered throughout the whole area.
Epigeic invertebrates were caught by pitfall traps (plastic pots) without preservative solution.In total, 60 traps were set in the forest and 40 traps were set at the clear-cut, in line 3 m apart.

RESULTS
During our research, 820 specimens of carabids belonging to 35 species were captured (Table I).The clear-cut was more abundant in both species (32) and the overall number of beetles (471, D = 57%); 349 specimens representing 25 species of carabid beetles were collected in the forest.Most specimens were collected in the spring (664 specimens, 29 species), while only 156 specimens of 21 species were collected in the autumn.The dominant species in the forest was Abax parallelus (D = 42%); Bembidion lampros dominated in the clear-cut (D = 25%).These two species together formed nearly 40% of the catch.The similarity between the forest and clear-cut communities was low (Jaccard index = 51%).Such a low similarity of populations is caused by a significant presence of forest species in the woodland (11 species, 44%) and open habitat specialists in the clearcut (15 species, 47%).
The epigeic activity of carabids in the forest was concentrated to evening and nighttime hours, with a peak period between 21:00 and midnight; in the clear-cut, activity had a distinctive daytime character, with peaks in the afternoon hours.
Spring epigeic activity was more or less evenly spaced throughout the entire day, declining late at night and in the early morning hours.Autumn was characterized by significant afternoon peaks of activity in a relatively narrow span, between 12:00 and 15:00 h, and a gradual decline during evening hours (Fig. 1).
A separate analysis of material collected in the spring and in the autumn showed that the factor lo-cality had greater significance in the spring.On the other hand, the factors of the daytime and temperature of the ground surface played a more significant role in the autumn (Fig. 3).
Samples collected in the forest growth were significantly influenced by season, while the influence of temperature was significantly greater in the clearcut (Fig. 4).Both localities differed in temperature regime, with temperatures more significantly fluctuating in the clear-cut than in the forest.While spring temperatures in the clear-cut fluctuated between 7 and 21°C, forest temperatures only ranged from 7 to 15°C.In the autumn, temperatures fluctuated between 1 and 21°C in the clear-cut, in the forest from 5 to 13°C.The amount of light at both localities was approximately the same (Fig. 4).Captured species were compared by their biotope preferences.The activity of forest species fluctuated during the year; they were active from evening to midnight in spring, and in the afternoon in autumn, respectively.Species in the clear-cut had the same pattern of activity, with afternoon peaks in both seasons; however, in the autumn, the peak was narrower       III.-50% of all specimens were collected in the period 12:00 -15:00 (Fig. 5).
For the second comparison, the carabids were sorted by reproduction biology.So-called spring breeders became activate after dusk in the forest, and during daylight in the clear-cut.Autumn breeders exhibited an elevated proportion of nocturnal activity; in the forest nocturnal activity clearly dominated, while in the clear-cut a second peak appeared -in addition to the afternoon maximum there was also increased activity in the early nighttime hours (Fig. 6).

Epigeic activity vs. exogenous factors
The influence of temperature and the daytime was tested on all species collected.The activity of 16 species was significantly affected by daytime, while 13 species were influenced by temperature (Table III).
In RDA analysis, with season, locality, temperature and daytime as the variables (Fig. 7), the majority of species were positively influenced by spring and the clear-cut environment.A bond to the forest environment was found in A. parallelipipedus, A. parallelus, Calosoma inquisitor and P. oblongopunctatus, which was the only species with higher autumnal activity.Only the dominant species significantly responding to at least one of the factors were used for the analysis.A total of 8 species responded significantly to daytime, while 9 species responded to temperature (Table II).
With the aid of generalized additive models (GAM), the influence of daytime and temperature abiotic factors on the epigeic activity of the dominant species were tested.Both Abax species and P. rufipes exhibited significant nocturnal activity, while typical clear-cut species A. similata, P. versicolor, P. cupreus, B. lampros together with P. oblongopunctatus (the only representative of forest specialist), showed striking diurnal activity (Figs.8a, 8b).
Increasing temperature of the soil surface positively influenced both Poecilus species as well as the other species with clear-cut preference: B. lampros, A. similata and P. melanarius.The forest species A. parallelus, A. parallelipipedus and P. oblongopunctatus preferred lower temperatures.

Diurnal activity vs. season
We studied the differences in epigeic activity during spring and autumn.Only P. cupreus provided a significant response to daytime in both periods, being active during the day, although this activity finished earlier in the autumn due to light deficiency.Four species -A.parallelipipedus, P. melanarius, P. versicolor, and P. cupreus -showed a significant connection between epigeic activity and temperature in both seasons (Figs.9a and 9b).P. melanarius and A. parallelipipedus preferred higher temperatures in the autumn compared to the spring.P. cupreus and P. versicolor were more active in autumn when the temperatures were even lower than in spring.

Diurnal activity vs. habitat
Three species (A.parallelipipedus, A. parallelus and A. similata) responded significantly to the daytime factor in both habitats (Figs.10a and 10b).Both species of Abax had a longer period of nocturnal activity in the forest than in the clear-cut.A. similata showed one significant peak at 15:00 in the forest and was active in a relatively narrow range of conditions, while in the clear-cut, the activity peak occured much sooner (around noon) and the range of activity was broader.The temperature factor was significant for A. parallelipipedus, A. parallelus and P. melanarius in both localities; however, while A. parallelipipedus showed a peak of activity under similar temperatures at both localities, A. parallelus was active in the clearcut in lower temperatures than in the forest and P. melanarius was active in the clear-cut in higher temperatures than in the forest (Figs.11a and 11b).

DISCUSSION
According to Park (1941), during their evolution animal populations tend to converge towards symmetry of activity in the framework of a 24-hour cycle.
In our study, the carabid population in a two-year old clear-cut represented the early stage of evolution, while the forest population represented the advanced (climax) stage.Thus, with the aid of Park's hypothesis we can predict that the activity in the open habitats will be probably asymmetric, while the activity in the forest would approach symmetry.If we include all the collected beetles in the analysis (beside the carabids, the most abundant were members of Staphylinidae and Geotrupidae), activity in the forest growth approached real symmetry (Fig. 2).In the case of carabid beetles, the clear-cut was significantly asymmetrical in diurnal activity, and the forest environment activity had a nocturnal character.Thus, our results do not correspond with those of Williams (1959), who confirmed the validity of Park's hypothesis for carabid beetles, but rather agree with Loreau (1986), who found predominately nocturnal activity in all dominant species of carabid beetles in a beech forest.
In our research, clear-cuts have greater species diversity than woodlands, which is again in agreement with many authors (Koivula, 2002, Magura, 2002, Ings and Hartley, 1999) who studied the influence of forest clearing on carabid beetle populations.
Carabids showed substantially greater activity in the spring (85% of the specimens), which is in agreement with Dondale et al. (1972), whose spring collection of Carabidae and Staphylinidae composed 85% of specimens collected per year.Novák (1980) reported a similar pattern in carabid catches depending on the season in the forest biotope.The significant influence of locality type for the catches in spring and autumn was given by the fundamental differences in the character of vegetation cover in open habitats between both seasons.In spring, the relatively thin herbal layer in the clear-cut was scattered with patches of exposed soil; in autumn the locality was covered with dense growth dominated by Impatiens glandulifera.Areas with dense vegeta-tion have a higher moisture level and therefore populations of invertebrates significantly differ from those in spring conditions (Haysom et al., 2003).
In spring, the activity of forest species reached a peak in the evening and early night hours, while in autumn activity had a peak in the afternoon.Ilosvay (1982) reports a significant peak of activity in the night hours (22:00-4:00) in June and July.Also, Thiele (1977) showed that most of forest species are night-active.In autumn, our forest species were mostly active in the afternoon hours, probably because of extremely low temperatures in the night and early morning.
Species in the clear-cut showed a peak of activity in the afternoon in both spring and autumn; the only difference was that the carabid beetles were active in a much narrower time range in autumn.We recorded a significant decline in activity in the autumn night and early morning hours.Similarly, Dondale et al. (1972) recorded the lowest number of collected specimens in the field in the period between 04:00 and 06:00.At this time, the temperature of the soil surface is the lowest, and according to Thiele (1977), temperature with light have the greatest influence on activity of carabids preferring open habitats.Preiszner and Karsai (1990), who conducted their research on sandy grasslands, reached this same conclusion.In our study, most of species (66%) showing a significant response to temperature were those living in open habitats.
Spring breeders in the forest environment were the most active at dusk and in the early night hours, while at the clear-cut their activity had a significant diurnal character.Novák (1980) recorded a predominance of activity in the daylight phase both in the forest and in the open habitats for spring breeders; but he also noted that the behavior of spring breeders is markedly influenced by microclimate.The evening and night peak of activity in the forest samples was connected with the presence of dominant A. parallelus, which is typically a moisturepreferring species (Thiele, 1964), and is therefore connected with nocturnal activity.On the other hand, we found predominantly diurnal activity in clear-cut carabids, which is in agreement with Kegel (1990), who studied the activity of carabid beetles in agrocenoses.
Autumn breeders showed a significant increase in activity in the night-time hours in comparison to spring breeders.In the forest environment noctur-nal activity predominated; in the clear-cut it was accompanied by activity in the afternoon hours.A predominance of nocturnal activity in autumn breeders was confirmed also by Kegel (1990) and Novák (1979Novák ( , 1980)), who recorded a peak of activity in the first two hours of night, which is in agreement with our results.P. melanarius and A. parallelipipedus show a preference for higher temperatures in the autumn than in the spring.This fact seems to be surprising with respect to the annual course of temperatures (autumn is substantially cooler).A preference toward higher temperatures in the autumn period is apparently related to the biology of reproduction of these two species -both are autumn breeders, which means that their sexual glands mature in autumn.The influence of temperature on egg size and the speed of development was studied in Notiophilus biguttatus (Ernsting and Isaaks, 1997), Amara spp.and Brachynus spp.(Saska andHonek, 2003a, 2003b).Higher temperatures accelerated the maturation of the sexual glands; this could provide an answer as to why the activity of the two otherwise nocturnal species mentioned above shifts to a diurnal pattern in autumn.They were probably actively searching for places with higher temperatures.
Discussion on selected species Thiele (1977) recorded A. parallelipipedus to be a eurytopic forest species without a pronounced relationship to temperature, but naturally preferring lower light intensities.Novák (1980) describes the peak of activity of this species in early night hours.In agreement with this, the activity of A. parallelipipedus was limited mainly by the daytime period (50% of specimens were collected in the interval 21:00 to 00:00), while in the autumn extremely low nighttime temperatures moved a limited activity to afternoon hours.
The activity of Abax parallelus showed the same pattern as in A. parallelipipedus.Thiele (1977) mentioned Abax parallelus as a stenotopic forest species; however, in our study A. parallelus was abundant in the material collected from the clear-cut; in autumn the numbers of specimens were equal at both localities.
Both species of Poecilus can be considered as typical representatives of open habitats, as they show a positive connection to the higher temperatures of the soil surface and to light.Thiele (1977) recorded them as euryphotic species.P. cupreus also showed noticeable diurnal activity (Novák, 1979(Novák, , 1980)), both in the forest and in open habitats.Our GAM showed that this species preferred the lower temperatures in the autumn than in the spring simply because the temperatures in autumn did not reach the level of the spring temperatures.P. cupreus in fact preferred the highest temperatures reached in both periods.
Pseudoophonus rufipes is regarded as a thermophile and xerophile field species with nocturnal activity (Thiele, 1977).According to Kegel (1990), this species in fields showed significantly higher activity during warm nights.While nocturnal activity dominated in our study, the number of specimens collected in the light phase was also relatively high (cca.33%).GAM analysis showed a significant dependence of epigeic activity of P. rufipes on the daytime, while dependence on temperature was not confirmed, despite the fact that its activity was divided more or less evenly throughout the entire day, with a peak in the evening and early nighttime hours.P. melanarius is regarded as a eurytopic and eurythermal field species preferring low light intensity (Thiele, 1977).In our research, its activity was significantly dependent on the temperature of the soil surface and it was active during periods with high temperatures.Alderweireldt and Desender (1990) also reported the diurnal activity of this species in the field with a substantial peak in the evening and early night hours.They assumed that declining light intensity increased its activity.This pattern would correspond to our results from the forest, while in the clear-cut the activity of this species was clearly diurnal.Novák (1979Novák ( , 1980) ) recorded an opposite pattern of activity for this species -diurnal activity in the forest and nocturnal in the open habitats.
Bembidion lampros is characteristic species of open habitats with a striking preference for diurnal activity in both field (Alderweireldt andDesender, 1990, Novák, 1979) and forest environments (Novák, 1980).In our research, this species was also significantly active during the afternoon and positively correlated with temperature.Thiele (1977) and Novák (1980) reported the species as a eurytopic forest one with a strong ecological bond to temperature and light; however, Loreau (1986) regarded this species as substantially nocturnal.This species dominated in the autumn and in the forest environment where its activity was significantly diurnal (15:00 -21:00) and was also influenced by temperature.

CONCLUSIONS
A total of 924 specimens representing 35 species of carabid beetles were collected.The carabid beetles were significantly more active in the clear-cut than in the forest and in the spring compared to autumn.
In the forest, nocturnal activity predominated, while in the clear-cut, activity had a significant diurnal character.This is contrary to Park's hypothesis on the development of populations toward symmetry of activity, which would predict symmetric activity in the forest environment representing the climax stage in the study.The time distribution of the epigeic activity of the entire beetle population better corresponds to Park's hypothesis.
In spring, activity was spread almost evenly throughout the day, while in autumn we recorded a significant afternoon peak of activity caused by low temperatures in the nighttime and early morning hours.
The daytime factor had a significant influence on 16 of the 35 species and the temperature of the soil surface significantly influenced 13 species.
CCA analysis showed a substantial increase in the explanatory potential of the factors of daytime and temperature for the activity of carabid beetles in the autumn compared to spring.CCA analysis also showed the substantial response of the activity of carabid beetles in the clear-cut to the soil surface temperature.
The activity of the forest species during the year shifted from the evening and night hours in spring to the afternoon hours in autumn, especially in the last weeks of the research when temperatures at night and in the early morning hours dropped below zero in the clear-cut.The activity of the species in the clear-cut maintained the same significantly diurnal pattern during the whole year.
The activity of spring breeders in the forest environment reached a peak at dusk, while their activity in the clear-cut was diurnal.Autumn breeders in both environments were mostly night-active.

Fig. 1 .
Fig. 1.Influence of daytime temperatures on the level of epigeic activities of carabid beetles in different seasons and at different localities.

Fig. 2 .
Fig. 2. Comparison of diurnal activities of carabid beetles and its dependence on the locality.

Fig. 5 .
Fig. 5. Comparison of the epigeic activities of forest species and open-habitat species of carabid beetles and its seasonal dependence.

Fig. 6 .
Fig. 6.Comparison of the epigeic activities of spring breeders and autumn breeders and its dependence on locality.

Fig. 8 .
Fig. 8. Response of epigeic activity of carabid species to (a) time and (b) temperature.For significance of response see TableII.
Fig. 8. Response of epigeic activity of carabid species to (a) time and (b) temperature.For significance of response see TableII.

Fig. 9 .
Fig. 9. Response of epigeic activity of carabid species to temperature in (a) spring and (b) autumn.For significance of response see Table III.
Fig. 9. Response of epigeic activity of carabid species to temperature in (a) spring and (b) autumn.For significance of response see Table III.

Fig. 10 .
Fig. 10.Response of epigeic activity of carabid species to daytime (a) in forest and (b) in clear-cut.For significance of response see TableIII.

Fig. 7 .
Fig. 7. RDA ordination illustrating distribution of the most numerous carabid species (dominancy more than 1.5%) in relation to environmental factors.Significance of factors: spring -F 40.85, forest -F 21.42, temperature -F 12.70, time -F 7.35, p for all factors lower than 0.01.

Fig. 11 .
Fig. 11.Response of epigeic activity of carabid species to temperature (a) in forest and (b) in clear-cut.For significance of response see Table III.a)b) . Canonical Correspondence Analysis and Generalized Additive Models for evaluating of results were created in the program CANOCO for Windows 4.5© (ter Braak and Šmilauer.1998); graphs were created in CanoDraw for Windows 4.0 and Microsoft Excel.

Table II :
Epigeic activities of the dominant species of carabid beetles and its dependence to predictor (daytime, temperature) (n.s.-nonsignificant, * p lower than 0.05, ** p lower than 0.01).

Table III :
Epigeic activities of carabid species in connection to predictor (daytime, temperature) and locality/season.